Setting The Record Straight
A good trade magazine holds both its news sources and its editors responsible for what they say

by Lee H. Goldberg

It's not unusual in this industry to see manufacturers making overly-optimistic claims about their products, and using subtle, often pointless, distinctions in their feature sets to declare their chip, box, or board an industry first. While we don't encourage such practices at analogZONE, it is our policy to publish the manufacturer's press release in its entirety as part of our product reviews and then pick apart any questionable claims they make in our commentary. In fact, taking regular jabs at those inflated claims has become one of our favorite sports. Of course, this also means that we expect to be held accountable for the statements we make.

This was exactly what happened in the case of my December 4, 2006 review of AMCC's new framer/pointer processors in networkZONE, where I'd cited Agere as an example of a company that appeared to have backed off their efforts in the SONET/SDH market.

Shortly after the review was posted I got a call from Nathaniel Grier, Senior Manager of the Network Connectivity Products group at Agere, expressing the company's great discomfort at my observation. In my defense I explained that I'd come to my conclusion honestly, based on reports from several sources that there had been a substantial cutback in their Networking Division's SONET products design staff, and my own observation that it had been a long time since they'd introduced a new framer, mapper, or other TDM device.

While he did not deny there had been some staff cutbacks, Mr. Grier made some good points that countered most of my concerns about the company's continued commitment to supporting SONET/SDH technologies. He pointed out that development efforts were continuing in several of Agere's design centers throughout the world and assured me that the company, in Grier's words, "is actively investing in feature enhancements and targeting new products for the L1 - L2 network connectivity space." I'm grateful for Agere's help in setting the record straight on this matter. I also welcome any other insights from readers which would provide additional perspectives on Agere's or any other company's strategic activities.

Mr. Grier also took me to task for not challenging AMCC's claim of several so-called firsts for their framer/mappers and for touting a number of features as unique to their product line. While he is correct that I did not take AMCC to task for some of their claims, I actually did do a little comparison shopping on Agere's and PMC-Sierra's web sites to find equivalent parts to gauge against AMCC's. Unfortunately, a side-by-side analysis was made more difficult by the company web sites (which were difficult to navigate) and the fact that I had to register to get any useful information. Being short on time, I relied on whatever product information I could glean from publicly-available documents and my own reviews of similar products to compare features, and thus may have missed several important points which might belie some of AMCC's statements in their December 4, 2006 product announcement.

In the interest of fairness to both Agere and our readers, I've turned over part of this Editorial to Agere's version of the facts. In his rebuttal to AMCC, Mr. Grier writes the following:

  1. AMCC is way off the mark with their claims of being the first to deliver this (chip's) capability and their claims of being best.
  2. Agere introduced the MARS10G Uni-PHY (T1610GA) solution two years ago and already has the product deployed and shipping in Tier 1 OEM systems. The Agere device integrates multirate CDR circuitry which supports 16 x OC-3, 16 x OC-12 and 4 x OC-48 with built-in auto-rate detection for all ports. Agere's mixed-signal expertise is further leveraged on the backplane side of the device to provide duplicated 4 x 2.5Gb/s (sic) serial backplane interfaces and duplicated 16x622Mb/s (sic) interfaces for legacy equipment support.
  3. The Agere TSOT1610G debuted over four (4) years ago and this technology was extensively leveraged in the T1610GA device -- a move which facilitated fast transition for our customers.
  4. Agere fully supports TMUX, DCS, ADM, UPSR, BLSR, and MSpring for 5G applications and further supports 10G cases too.
  5. More importantly, the T1610GA family of devices with our secondary line interface, duplicated backplane interfaces and hitless protection switching support DWDM and TCM applications, which the AMCC parts apparently do not support.
  6. The Agere solution implements 20G of switching capability and also provides overhead transparency, and we have comparable power dissipation for the applications described by AMCC.
  7. With no power advantage, no feature advantage and availability which won't occur until 2Q07, we're happy to see the announcement from AMCC on their Drava and Mura devices. It further confirms Agere's leadership in the Telecom Network Connectivity space.

Below are three links which provide an overview of our current generation MARS T-PHY solution. The device, named TSOT1610GA, has the "A" extension because of the added analog (really mixed-signal) capability delivered by the multirate CDR. Note also that the release is not a product pre-announcement -- we specifically avoided pre-announcing these components.

I think that Agere's commentary is helpful in making a more accurate comparison between Agere's and AMCC's devices. It also reinforces the point I briefly made that Agere has had a similar part (the TSOT1610G mentioned above) available for some time. A bit further on in my review, I also noted that the primary distinguishing feature that AMCC enjoyed over the older Agere device was more fully user-configurable overhead transparency capabilities -- a point which Agere disputes.

Once again, time pressure to get this week's issue out has not permitted me to do all the legwork necessary to finish a detailed follow-up comparison of each device's overhead processing capabilities, for a definitive answer about who boasts the best feature set. I'll try to get back to you with some more hard facts about whose overhead processing scheme has the edge, but in the meanwhile I'd welcome any insights that you, Agere, or AMCC might have on this, or any other aspect of SONET/SDH technology, on analogBLOG, linked at the bottom of this page.

And, as always, you can also send your questions and comments, or hold me responsible for anything I write, at: lgoldberg@green-electronics.com

PS. If you're one of our many sharp-eyed readers, you may have noticed that while I'm posting here a public apology to Agere for my oversights in networkZONE, this same issue carries a semi-glowing review of another Agere product in our connectivityZONE. Suspicious as this might seem, it was not a premeditated move, but simply one of those weird synchronous events that occur from time to time. Of course, there is no way I can prove to you that there was no calculation behind this...but regardless of coincidence, appearance, or who we manage to tick off, you can be sure that we'll continue to give you the most accurate information we can, tell things as we see them, and bite the hand that feeds us.

analogZONE
(c) 2006. All rights reserved.